Introduction
Pellet extrusion unlocks material freedom that filament printing cannot: custom polymers, blends, recycled materials, and experimental compounds. At the same time, many users find that early prints can look rough, inconsistent, or straight-up unusable.
This page explains why pellet-extruded prints can look bad, and more importantly, how to make them better. The explanations apply broadly to most desktop screw-driven pellet extruders. Where helpful, we use Materium as a reference example of a precision-focused design—not as the only solution, but as a clear illustration of what careful extrusion design and tuning can achieve.
If your pellet prints show stringing, rough surfaces, gaps, or unstable flow, this guide is meant to help you understand what’s happening and what to try next.
(image)
Why Pellet Prints Looks Bad
Most print-quality issues in pellet extrusion compounds on top of another and tuning them can be an endless goose chase. These are common across many machines.
1. Pellet Quality
Pellets are not as standardized as filament. When compared in concept, the only thing that’s fundamentally different is how filaments have uniform diameter, uniform blend, and most importantly : optimized exclusively for FDM. These optimizations such as plasticizers, flow enhancer, melt temperature reducer, ultrafine fillers, extra crosslinker, among many other things can easily be taken for granted on how specific they are to FDM filaments.
In contrast, polymer pellets (also called resin) are raw, pure polymers unless stated otherwise. Most of these are for injection molding, blow molding, pressure cast, or extruded into rods. For most cases, none of their intended use-case cares about the one thing 3D printing needs the most : precision melt deposition with minimal shrinkage. The title of this section may be misinterpreted; high quality pellets for injection molding has almost nothing to do with how good they are when 3D-printed.
Ironically, a ‘high quality’ injection pellets has some of the worst qualities for FDM 3D Printing. Let’s address basics first without going into filament perspective yet, which are usually “not dry”, “tangled”, “old”, “bad batch”, “brittle”, or “inconsistent diameter” because bad news, these problems exists for pellets too. But we will separate that. First, These high-level quality issues are :
- ‘High purity’ means small debris might exist. Meanwhile, FDM printing with 0.4mm nozzle requires perfect purity. Injection mold nozzles are much bigger than FDM, sometimes 5mm and up. This is why a speck of rock inside a single pellet can still qualify as high quality. But 0,4mm nozzles are the domain of exclusively FDM. Even by mishandling, (some rocky dust got into the pellet bucket) clogging is only a matter of time.
- High molecular weight means the injection-molded product will feel ‘heavy’ or ‘solid’, even ‘premium’. For FDM this means massive warping. Injection molds just push with tonnes of pressure until this effect becomes manageable. FDM extrudes high-MW polymers into free air. Note : some high-MW polymers do not shrink as much as high-MW versions of PP.
- High Foaming/Expanding Agents. These are used for making cushioning objects like footwear soles. Injection molding doesn’t make infills like FDM does, and having foaming agents makes micro-structures, just like infills. These additives can make the polymer expand to 10x its original size. Note that foaming pellets are injected into closed molds with very high pressure to force them to conform to shape. When extruded into free air like FDM, the extrudate is basically incomprehensible.
Other ‘high quality’ implications that doesnt matter for 3D printing :
- Colors : nice color spread, deep colors
- High Flowability : extrudes more with less pressure
Quality implications that directly impact 3D printing it :
- MFI (empirically found that anything lower than 5 grams/10mins is unusable with Materium with 0.6mm nozzle). See : Melt Flow Index
- Mechanical properties (Tensile Modulus, Tensile Strength, Charpy/Izod Impact, Flexural Strength, Flexural Modulus, only if you care. We are way past “is it strong?” here)
- Weldability (layer adhesion; implies low surface tension as a melt)
- Melting point and Softening Point (for setting extrusion temperature and drying temperature)
- Virgin or Regrind (industrial term for recycled; roughly implies molecular weight, more recycled usually means lower MW)
Quality implications that impacts but never mentioned :
- Pellet size and shape (smaller and more spherical is better). See : CHOPPER
- Pellet uniformity (varying pellet size means varying volumetric output)
- Shrinkage percentage (affects how prone to warping)
- Processing temperature (most supplier rely on buyers to telepathically know what temps to use. Luckily, filament temps are a good start here)
- Chemical variants of the same polymer (most low-price suppliers labels everything that ends with ‘-PE’ as ‘PE’ regardless of whether it’s actually LDPE/HDPE/LLDPE/HMWPE/UHMWPE. If this is not true in your experience, bless you. In developing areas like SEA countries, this is more pronounced and doing meaningful research becomes difficult).
Even a well-designed extruder cannot compensate for very poor pellet input. The above helps to calibrate what good or bad pellets are, before prematurely judging them with filament knowledge.
2. Inconsistent Melting
Unlike filament, pellets must be:
- Captured (in the feed zone)
- Compressed (drags against the screw and the barrel at the same time, until hot and becomes soft)
- Melted
- Metered (extruded with precision)
If melting is incomplete or uneven:
- Unmelted fragments can reach the nozzle
- Flow becomes pulsating instead of smooth. See : Interpreting Extrudates
- Retraction fails
- Layer adhesion fails
- Wall quality suffers
This is often caused by insufficient compression, off-tuned PID, bad insulation, or running the extruder too fast/slow for the material.
3. Temperature Management Issues
Temperature problems are rarely just “too hot” or “too cold.” More often, they are uneven, or abrupt.
Common symptoms:
- Too hot → popping bubbles, dripping, sagging, degraded material
- Too cold → poor layer adhesion, lowered flow, snapping lines, skipping motor, wrong line width. See : Temperature and Flow Tuning
- Hot feed zone → pellets soften too early and backflows. See : Understanding Screw Extrusion
- Cold melt zone → does not extrude, or extruding but motor works too close to skipping
Pellet extruders are far more sensitive to temperature distribution than filament hotends.
4. Nozzle and Die Effects
The nozzle is not just an exit hole—it controls pressure and flow stability.
Issues include:
- Sharp internal corners → turbulent flow
- Poor land geometry → pressure loss and inconsistency. See : Adjusted Terminology and Understanding Screw Extrusion
- Very small nozzles on unstable melt → visible defects
Large nozzles often hide problems; smaller nozzles expose them.
5. Bridging
Bridging is one of the simplest issue to deal with; although it may not be easy. When seen as liquid molecules, the setups we use to transport pellets at this scale (tubes, hoppers, channels) are microscopic capillaries compared to the molecule size (pellet size). This means sometimes the pellet flow gets jammed for no apparent reason that is easily restored by disturbing the jam/clog by poking them a little.
In injection molding, this doesn’t usually bother us much and solutions already exist, such as embedding stirrers or vibrators. In 3D Printing, much higher stakes. Bridging can happen at any point between the hopper (pellet bucket) to the melt zone. This eventually causes ‘starvation’ (See : Understanding Screw Extrusion and Hopper Setup) and is equivalent to filament grinding.
Bridging is caused by one or two reasons :
- pellet path too narrow (make tubes bigger)
- pellet path too long (make shorter, or attach hopper directly on extruder)
- pellet tube (if using) has too much friction or grooves (stuck pellets builds up)
6. Slicer Settings
This could be the easiest to tune rather than other issues above, and the most important setting if Flowrate / Multiplier.
A 5% difference in Multiplier looks like this : (MIRA)